This is not a question of ideology or personal beliefs. It’s just a matter of fact that AI systems, especially LLMs, are not capable of producing reliable results.

This is a fundamental issue and for many uses we still need the old school way of computing that produces reliable results. It doesn’t happen accidentally, you know, a lot of resources went into this. AI just isn’t there yet and probably won’t be there anytime soon, no matter how much compute we throw at it.